The goal of this project is to elucidate structure-function relationships in macromolecular machines. During FY12, we studied: chaperone-assisted proteases involved in protein quality control and cell regulation; membrane remodeling; and two components of pathogenic bacteria. (1) All cells must be capable of degrading aberrant and foreign proteins that would otherwise pollute them. Programmed degradation of regulatory factors also contributes to controlling the cell cycle and to generating peptides for immune presentation. These activities are all carried out by energy-dependent proteolytic machines, which generically consist of two subcomplexes - a protease and an ATPase/unfoldase. Since 1995, we have studied the Clp complexes of E. coli, considered as a model system. We showed that the protease ClpP consists of two apposed heptameric rings and the cognate ATPase - either ClpA or ClpX - is a single hexameric ring. ClpA/X stack axially on one or both faces of ClpP s. We went on to show that substrate proteins bind to distal sites on the ATPase and are then unfolded and translocated axially into the digestion chamber of ClpP. These studies all dealt with homomeric ring complexes. ClpP in plants depart from this paradigm in being heteromeric. Chloroplast ClpPs associate multiple different subunits. In FY12, we completed our structural analysis of ClpP from the green alga Chlamydomonas reinhardtii (1). As with land plants, this ClpP has both active subunits (3 ClpPs) and inactive subunits (5 ClpRs). It also has two ClpT subunits which, like land plant ClpTs, show Clp-N domains. ClpTs are believed to function in substrate binding and/or assembly of the two heptameric rings. Negative staining electron microscopy showed that the Chlamydomonas ClpP complex retains the barrel-like shape of homomeric ClpPs, with 4 additional peripheral masses which may represent either the additional IS1 domain of ClpP1 (a feature unique to algae), or ClpTs, or extensions of ClpR subunits. (2) Membrane Remodeling. Remodeling, a process in which lipid bilayer structures are reconfigured by interacting proteins, is central to the functioning and metabolism of cells. We are investigating this phenomenon by using cryo-electron microscopy to characterize the effects of remodeling proteins on large lipid vesicles in vitro. Endophilin A1 is a BAR (Bin/Amphiphysin /Rvs) protein abundant in neural synapses that senses and induces membrane curvature, contributing to neck formation in pre-synaptic vesicles. We found that, on exposure to endophilin, vesicles convert rapidly to coated tubules whose morphology reflects the local concentration of endophilin. Their diameters and curvature resemble those of synaptic vesicles in situ. 3D reconstructions of quasi-cylindrical tubes revealed arrays of BAR dimers, flanked by densities that we equate with amphipathic. We also observed the compression of bulbous coated tubes into 7nm-wide cylindrical micelles, which appear to mimic the penultimate (hemi-fission) stage of endocytosis. Starting in FY11, this project was extended to alpha-synuclein (aS). Natively unfolded in solution, aS accumulates as amyloid in neurological tissue in Parkinsons disease, and also interacts with membranes under both physiological and pathological conditions. We used cryo-EM in conjunction with electron paramagnetic resonance (EPR) and other techniques to characterize the effect of aS on lipid vesicles. The products obtained depend on the protein : lipid ratio. At a molar ratio of < 1: 40, POPG vesicles are converted into cylindrical micelles 5nm in diameter, together with bilayer tubes of various widths (15 - 50 nm ). Between 1 : 20 and 1 : 10, cylindrical micelles are produced exclusively. Other negatively charged lipids (DMPG, DLPG, DAPG) exhibit generally similar behavior. At higher protein: lipid ratios (> 1: 10), cylindrical micelles are replaced by discoid particles, 7 - 10 nm across. In FY12, these studies were completed and have been accepted for publication (5). 3) The resistance of pathogenic bacteria to conventional antibiotic drugs is a growing problem in combating human infections. An alternative therapeutic strategy uses virus-derived proteins, called lysins, to kill Gram-positive bacteria by damaging the peptidoglycan layer in their envelopes. However, a major obstacle to this approach is that lysins cannot cross the protective outer membrane. In this study, a chimeric lysin was produced and shown to kill Gram-negative bacteria. A crystal structure was obtained for a Yersinia pestis toxin called pesticin and used to guide the design of a hybrid lysin in which a domain from pesticin was fused withT4 lysozyme, an archetypal lysin. This hybrid protein was shown to cross the outer membrane and kill model E. coli cells as well as Yersinia strains. Cryo-EM of whole bacterial cells treated with the lysin showed extensive disruption of their membranes (4). Importantly, killing only affects cells that produce the membrane protein FyuA4, a virulence factor, and can therefore be used against essentially any pathogenic strain that expresses FyuA4. 4) Neisseria are Gram-negative bacteria of which two pathogenic species invade the human urogenital tract and nasopharynx, causing gonorrhea, meningitis and other systemic infections. Neisseria require iron for survival and virulence. While most Gram-negative bacteria secrete siderophores to scavenge iron from the environment, Neisseria extracts iron directly from human sources such as transferrin. The Neisserial transferrin transport system consists of two large surface proteins: transferrin binding protein A (TbpA), a 100 kDa integral outer membrane protein, and TbpB, an 80 kDa transferrin receptor. TbpA binds apo and holo transferrin, whereas TbpB associates only with the iron-bound form of transferrin. In order to learn how TbpA and TbpB interact to bind human transferrin and extract its tightly bound iron at physiological pH, a combined structural approach was used. A crystal structure determined for the TbpA-(apo)hTf dimer and a SAXS structure for the TbpB-(holo)hTf complex led to a model for the TbpA-TbpB-(holo)hTf triple complex. We tested this model - with a positive outcome - by negative staining electron microscopy supplemented by computational class-averaging (6). Our results suggest that Neisseria cannot utilize the whole serum transferrin iron supply and the primary function of TbpB is to select and concentrate on the cell surface only those forms of transferrin that can be taken up.